Name
Abstract | ||
|---|---|---|
We present a Bayesian approach to adapting parameters of a well-trained context-dependent, deep-neural-network, hidden Markov model (CD-DNN-HMM) to improve automatic speech recognition performance. Given an abundance of DNN parameters but with only a limited amount of data, the effectiveness of the adapted DNN model can often be compromised. We formulate maximum a posteriori (MAP) adaptation of parameters of a specially designed CD-DNN-HMM with an augmented linear hidden networks connected to the output tied states, or senones, and compare it to feature space MAP linear regression previously proposed. Experimental evidences on the 20,000-word open vocabulary Wall Street Journal task demonstrate the feasibility of the proposed framework. In supervised adaptation, the proposed MAP adaptation approach provides more than 10% relative error reduction and consistently outperforms the conventional transformation based methods. Furthermore, we present an initial attempt to generate hierarchical priors to improve adaptation efficiency and effectiveness with limited adaptation data by exploiting similarities among senones. |
Year | Venue | Keywords |
|---|---|---|
2015 | 16TH ANNUAL CONFERENCE OF THE INTERNATIONAL SPEECH COMMUNICATION ASSOCIATION (INTERSPEECH 2015), VOLS 1-5 | deep neural networks, hidden Markov model, Bayesian adaptation, automatic speech recognition |
Field | DocType | Volume |
Feature vector,Pattern recognition,Computer science,Artificial intelligence,Maximum a posteriori estimation,Prior probability,Hidden Markov model,Vocabulary,Machine learning,Approximation error,Linear regression,Bayesian probability | Journal | abs/1503.02108 |
Citations | PageRank | References |
8 | 0.43 | 24 |
Authors | ||
6 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Zhen Huang | 1 | 100 | 11.60 |
| Sabato Marco Siniscalchi | 2 | 310 | 30.21 |
| I-Fan Chen | 3 | 110 | 10.72 |
| Jinyu Li | 4 | 915 | 72.84 |
| Jiadong Wu | 5 | 12 | 1.18 |
| Chin-Hui Lee | 6 | 6101 | 852.71 |